Remove pt cut to SV reconstruction of dileptons for DiMuonSoft line
During the analysis of Ks2mu using 2024 data we realized that the reconstruction sequence used in HLT1 was making us lose about 33% efficiency due to the reconstruction of the SV. There was a pt cut of 200 MeV, on the first figure we see this in the pt distribution of the muons. As far as we know this cut did not exist in Run 2.
We have removed this cut (for now) only for our desired line HLT1DiMuonSoft.
We used the HLTEfficiency checker to compute the rate before and after the changes. This was done for the sequence hlt1_pp_forward_then_matching_and_downstream_with_parkf_1200kHz_M23 and the input data 2024_hlt1_mdf_dump_291820. We ran for 100k and the results were the following
Before: Hlt1DiMuonSoftDecision: Incl: 4.500 +/- 1.162 kHz, Excl: 3.000 +/- 0.949 kHz
After: Hlt1DiMuonSoftDecision: Incl: 5.400 +/- 1.273 kHz, Excl: 3.000 +/- 0.949 kHz
Both values are the same within uncertainties.
Using simulations we computed the signal efficiency using 63k events and we saw an increase of about 33% from 1.21% to 1.79%
The throughput was computed in the git pipeline giving the following results:
For sequence hlt1_pp_forward_then_matching: % change -2.13%
For sequence hlt1_pp_default: % change -0.58%
Attach screenshots of the throughput from the git pipeline
Attach the monitoring pt plot from Allen after the changes.
To do:
- Run over 2025 data
